Measurement of Metabolites asIndicators of Exposure toChemicals

STUART A. SLORACH

1 INTRODUCTION

Many chemicals to which animals or plants are exposed undergo metabolismprior to storage in, or excretion by, the organism. In some cases, measurementof the levels of metabolites in tissues or excretion products can be used asan indicator of exposure to such chemicals. Measurement of metabolites asa means of monitoring exposure may be especially valuable when practicaldifficulties exist in determining exposure to a specific chemical. Such asituation can arise either when a chemical is metabolised very rapidly orwhen it is difficult to assess exposure, especially long-term exposure, byenvironmental monitoring, e.g., when dosing is by inhalation during whichair concentrations of a chemical fluctuate widely. Assessing long-termexposure to illustrative air pollutants, such as passive smoking or industrialexposure to volatile organic compounds, exemplifies an area where metabolitemeasurements may provide an accurate indication of exposure. However,for metabolite measurement to be useful as an indicator of exposure tochemicals, several criteria should be fulfilled:

(1) A metabolite should be specific for the chemical concerned, unlessisotope studies are contemplated. By contrast, metabolite determination islikely to be of little value if the chemical is metabolised completely to "non-specific" metabolites, such as water, carbon dioxide, acetate, chloride, orurea.

(2) The metabolism of a chemical in the organism of interest shouldbe known. Studies on the metabolism of drugs and other xenobiotics haveshown that for many compounds there are large inter- and intra-speciesdifferences in metabolism. There may also be differences in metabolism inthe same individual, depending on age, and hormonal and nutritional status.All these factors complicate the use of metabolite determinations in assessingexposure to chemicals.

(3) As with all monitoring, reliable methods of sampling and analysismust be used and analytical quality assurance must be carried out to obtainvalid results.

Methodsfor AssessingExposureof Human and Non-Human Biota. Edited byR.G.TardiffandB.Goldstein@ SCOPE 1991. Published by John Wiley & Sons LId

324 METHODS FOR ASSESSING EXPOSURE OF HUMAN AND NON-HUMAN BIOTA

A wide range of tissues, body fluids, and excretion products are potentiallyuseful for the determination of metabolites in vivo: for example, (a) bloodor fractions thereof, saliva, and milk; (b) adipose tissue, cerumen, teeth,hair, and nails; and (c) urine, feces, and expired air. A much wider rangeof tissues can be used in studies of lifetime exposure to chemicals whichaccumulate in the body, if tissues can be collected from cadavers at autopsy.

Metals present in tissues or body fluids (e.g., lead in blood, bone orteeth, cadmium in kidney cortex, and mercury in hair) may be in a formother than that in which they were introduced into the body, i.e., they maybe present as metabolites. However, relatively little is known about theactual chemical form in which these metals are present in the tissues concernedand, for exposure monitoring purposes, they are usually determined asthe metal, and not as metabolites, and thus they will not be discussed here.

Examples are given below of how determination of metabolites can beused to assess exposure to a wide range of chemicals, including nicotine (asan indicator of environmental exposure to tobacco smoke), industrial solvents(e.g., benzene, m-xylene, and styrene) and persistent pesticides (e.g., p,p'-DOT).

2 COTININE IN URINE AS A MEASURE OFENVIRONMENTAL EXPOSURE TO NICOTINE FROMTOBACCO SMOKE

Cotinine is the major metabolite of nicotine found in the urine. Themeasurement of cotinine has advantages over determination of the parentcompound, nicotine, because the circulating half-life of cotinine in adults is19 to 40 hours (Langone et al., 1973; Benowitz et al., 1983), compared toless than 30 minutes to 110 minutes for nicotine (Isaac and Rand, 1972;Rosenberg et al., 1980). Cotinine can be measured at concentrations of lessthan 500 pg per millilitre by a sensitive radioimmunoassay (Haley et at.,1983; Hill and Marquardt, 1980). Thus, the presence of cotinine can beused as an indicator of chronic exposure to tobacco smoke.

Greenberg et al. (1984) measured the concentrations of nicotine andcotinine in the saliva and urine of 32 infants with household exposure totobacco smoke and 19 unexposed infants. The concentrations were signifi-cantly higher in the exposed group, the best indicator of exposure beingurinary cotinine/creatinine ratio (median in the exposed group 351 ng permg, median in the unexposed group 4 ng per mg (P<0.0001)). There wasa direct relation between cotinine excretion by infants and the self-reportedsmoking behaviour of the mothers during the previous 24 hours (r=0.67,P<0.0001). Thus, the results indicated that infants exposed to tobaccosmoke absorbed nicotine from it and that urinary excretion of cotinine ininfants is a relatable measure of such exposure.

METABOLITES AS INDICATORS OF EXPOSURE TO CHEMICALS 325

Jarvis et al. (1984) studied the relationship between several biochemicalmarkers of smoke absorption and self-reported exposure to passive smoking.One hundred nonsmoking patients attending hospital outpatient clinicsreported their degree of passive exposure to tobacco smoke over thepreceding three days and provided samples of blood, expired air, saliva,and urine. Although the absolute levels were low, the concentrations ofcotinine in all body compartments surveyed were systematically related toself-reported exposure. Salivary nicotine concentration also showed a linearincrease with degree of reported exposure, although this measure wassensitive only to exposure on the day of testing. Carbon monoxide,thiocyanate, and plasma nicotine concentrations were unrelated toexposure.

Data from these studies indicated that cotinine excretion in the urine is

a positive indicator of nicotine exposure and potentially of passivesmoke exposure. Urine samples are particularly suited to epidemiologicalinvestigations, since they can be obtained from all age groups, even infants.

3 PHENOL IN URINE AS A MEASURE OF BENZENEEXPOSURE

Benzene is metabolised principally into phenols, which are excreted in theurine as glucuronide and sulphate conjugates. Measurement of phenol inurine has been routinely used earlier for assessing benzene exposure inindustry. This is based on the assumption that between 20 and 40 percentof benzene in blood is metabolised to phenol. However, phenol in urine isnot a specific indicator of exposure to benzene and the excretion of phenolin urine can be increased by diet, hepatic disturbances, intestinal putrefactiveprocesses, and drugs such as aspirin (Fishbeck et al., 1975). The finding ofphenol levels above 20 mg!l suggests exposure to benzene, if othercircumstances causing increased phenol excretion have been excluded(Lauwerys, 1979). Phenol levels in urine can be determined by gaschromatography (van Haaften and Sie, 1965), and can thus monitor heavyexposures to benzene. However, sensitive methods are now available forthe determination of benzene in blood. Such methods are more reliablethan phenol tests for assessing both low level exposure and uptake ofbenzene, since low levels of benzene can be detected in blood even whenphenol levels in urine are regarded as normal (10 mg!l or less) (Braier etal., 1981).

The development of a sensitive technique for the isolation and quantifi-cation of trans-muconic acid, an open-chain urinary metabolite of benzene,

has provided a further method to monitor exposure to benzene (Gad-ElKarim et at., 1985).

326 METHODS FOR ASSESSING EXPOSURE OF HUMAN AND NON-HUMAN BIOTA

4 m-METHYLHIPPURIC ACID IN URINE AS ANINDICATOR OF ABSORBED m-XYLENE

m-Xylene is the principal component of commercial xylene mixtures widelyused as solvents in paints and thinners. Xylene isomers are highly solublein blood and tissues and undergo metabolism to methylbenzoic acids, whichare then conjugated to glycine to yield methylhippuric acids. Thus m-xyleneyields m-methylhippuric acid which is excreted in the urine.

Biological monitoring of occupational exposure to xylene can be carriedout by measuring xylene levels in blood and analysis of the metabolites inurine. Measurement of m-methylhippuric acid has been proposed (Ogataet al., 1970; Mikulski et al., 1972) and the method has been tested underexperimental conditions by Sedivec and Flek (1976).

Methylhippuric acids in urine can be determined by gas chromatographyafter alkaline hydrolysis (Engstrom et aI., 1976).

Engstrom et al. (1978) have evaluated different methods to assessoccupational exposure to xylene. They concluded that the determination ofmethylhippuric acids in urine samples collected at the end of the workingday could be used for evaluating the average xylene exposure during thepreceding day. The concentration of methylhippuric acid in a morningsample at the end of the work week provides a crude estimate of the xyleneexposure during the preceding week.

Senczuk and Orlowski (1978) investigated the absorption of m-xylene andthe excretion of m-methylhippuric acid in the urine in 10 volunteers. Theyfound a relationship between the absorbed dose of m-xylene and the rateof excretion of m-methylhippuric acid between the sixth and eighth hour ofexposure (correlation coefficient = 0.99). They proposed that exposure toxylene be monitored by determining

(1) the concentration of m-methylhippuric acid in urine during the 6thto 8th hours of daily exposure or during the first four hours after the endof exposure (8th to 12th hours); and

(2) the amount of the metabolite in urine collected during the last twohours of exposure.

From these data, it is possible to estimate not only the probableconcentration of m-xylene during exposure but also the amount absorbed.

5 MANDELIC ACID IN URINE AS AN INDEX OFSTYRENE EXPOSURE

In humans, styrene is metabolised mainly in the liver, first to styreneoxide, which is biologically active and can bind covalently to cellularmacromolecules. Styrene oxide is hydrated to phenylethylglycol and can beconjugated with glutathione. Phenylethylglycol can be conjugated with

METABOLITES AS INDICATORS OF EXPOSURE TO CHEMICALS 327

glucuronic acid and excreted in the urine or oxidised to mandelic acid (MA),which is partly oxidised further to phenylglyoxylic acid (PGA). A smallfraction of styrene can be metabolised to vinyl phenol and phenylethanol.Only small amounts of absorbed styrene are excreted unchanged in exhaledair; most of it is excreted as metabolites into the urine. The two mainmetabolites, MA and PGA, correspond to about 92 percent of the retaineddose of styrene, of which approximately 34 percent is excreted during an 8-hour period of exposure. The ratio of the two metabolites (MAlPGA) isabout 3:1 at the end of the exposure, dropping to 4:5 approximately 14hours after cessation of exposure (Guillemin and Bauer, 1979).

Determination of MA and/or PGA in urine can be used to assess exposureto styrene. Several analytical methods are available, but according toEngstrom (1984) calorimetric determination of MA and PGA in urine issubject to interference and gives a relatively high background level, makingthe method less suitable for the exact estimation of MA and PGA excretionat low doses of styrene. The relative error decreases towards higherconcentrations, as the unspecific interference is relatively constant and notaffected by the styrene load. With more specific methods, such as gas andliquid chromatography, the normal values are so low that there is no problemwith the variability in normal levels. Exceptions are consumption of somedrugs, which may be metabolised to MA, and exposure to ethylbenzenewhich is excreted in the urine as MA. A greater analytical problem is theinstability of samples. In urine, PGA and MA are both prone to changeduring storage at room temperature as well as at +6° C. A low pH seemsto improve the stability.

Droz and Guillemin (1983) described a mathematical model to developthe best strategy for biological monitoring of exposed workers. In theircomparison of biological indicators of styrene exposure (MA, PGA, MA +PGA), while taking into account pharmacokinetics, analytical variance, fieldand laboratory practicability and inter- and intra-individual variance, MAmeasurement in urine had the highest rating, with respect to measurementsboth at the end of shifts and the day after exposure.

Engstrom (1984) recommended the determination of MA in urine (wholeshift samples) using specific analytical methods for the biological monitoringof exposure to styrene. Such results should be corrected for the creatininecontent of the urine, and storage of samples before analysis should beavoided. In assessing results, the type of exposure, the physical work load,and individual factors (such as alcohol intake) should be taken into account.

6 :PDE IN ADIPOSE TISSUE, MILK FAT, OR CERUMENAS AN INDEX OF EXPOSURE TO DDT

The extensive,and in somecountriescontinuing,use of largequantitiesofthe persistent insecticidep ,p'-DDT has led to widespread contamination of

328 METHODS FOR ASSESSING EXPOSURE OF HUMAN AND NON-HUMAN BIOTA

the environment.One resultof this is that levelsin humanmilk in somecountries are so high that the intake by breast-fed infants may exceed theAcceptable Daily Intake recommendedby the World Health Organisation.However, there is no evidence that such an intake has a deleterious effecton the health of infants.

The main metabolite of p,p'-DDT in man is p,p'-DDE, which is retainedin in adipose tissueof the body. The levels of p,p'-DDE in adiposetissue,breast milk fat, blood (or fractions thereof), or cerumen can be used as anindicator of exposure to p,p'-DDT.

A large number of gas chromatographic methods is available for theanalysis of p,p' -DDE and related lipid-soluble, persistent organochlorinecompounds (Slorach and Vaz, 1983).

Studies of the levels of p,p'-DDE and related compounds in breast milkhave been carried out since the 1950s and thus there is a considerable bodyof data to enable the determination of time-trends. However, for a numberof reasons the interpretation of the results of studies on the levels of p,p'-DDE in adipose tissue or breast milk fat is difficult. First, unless the personconcerned is exposed directly to p,p'-DDT, as in the case of occupationalexposure, it is likely that there will be mixed exposure to both p,p' -D DTand p,p'-DDE via foodstuffs of animal origin (e.g., fish, meat, and dairyproducts). In such cases, the level of p,p'-DDE in human tissues does notreflect exposure to p,p'-DDT alone but also includes exposure to preformedp,p'-DDE. In countries where p,p'-DDT is still used in agriculture andvector control programmes, the ratio p,p'-DDE/p,p'-DDT levels in breastmilk fat is very much lower than in countries that have stopped using DDT(Slorach and Vaz, 1983; Jensen, 1983). The explanation for this is that inthe latter countries exposure is mainly via foodstuffs which contain mainlyp,p'-DDE and little or no p,p'-DDT; whereas, in the former exposure top,p'-DDT may occur both via foodstuffs and via other routes such asinhalation and skin contact.

When p,p'-DDE levels in human milk fat were used as an indication ofexposure to p,p' -DDT, in the same individual those levels can be affectedby factors such as parity and length of previous lactation periods and thetime post-partum at which the samples are collected.

Levels in adipose tissue or breast milk fat can also be used as an indexof exposure to other persistent organochlorine pesticides. For example,exposure to aldrin can be monitored by measuring the levels of its metabolitedieldrin. Likewise, exposure to heptachlor can be assessed by measuringlevels of heptachlor epoxide in adipose tissue and breast milk fat.

Readers interested in the use of breast milk for biological monitoring arereferred to the excellent review by Jensen (1983).

METABOLITES AS INDICATORS OF EXPOSURE TO CHEMICALS 329

7 CONCLUSIONS

As the above examples show, the determination of metabolites in bodyfluids and tissues can provide an indication of exposure to certain chemicals.This method is especially useful in cases where metabolism yields a metabolitewhich is specific and stable. Unfortunately, the interpretation of the resultsof this type of biological monitoring of exposure to chemicals may becomplicated by the facts that there are intra-species and intra-individualdifferences in the metabolism of some chemicals and that some metabolites

are not specific to a single chemical. The development of better analyticalmethods and advances in our knowledge of the metabolism of environmentalchemicals will provide a firmer foundation for this type of biologicalmonitoring in the future.

For further information on the use of metabolites in biological monitoringthe reader is referred to Sheldon et al. (1986); Carson et al. (1986); Linch(1974); Aitio et al. (1984); Patty's Industrial Hygiene and Toxicology (Cralley,1985); and a publication from the American Conference of GovernmentalIndustrial Hygienists (ACGIH, 1985).

8 REFERENCES

ACGIH (1985). Threshold Limit Values and Biological Exposure Indices for/985-1986. American Conference of Governmental Industrial Hygienists, Cincin-nati, Ohio.

Aitio, A., Riihimaki, V., and Vainio, H. (1984). Biological Monitoring andSurveillance of Workers Exposedto Chemicals. Hemisphere, Washington, D.C.

Benowitz, N.L., Kuyt, F., Jacob, P., Jones, R.T., and Osman, A. (1983). Cotininedisposition and effects. Clin. Pharmacol. Ther. 34, 604-611.

Braier, L., Levy, A., Dror, K., and Pardo, A. (1981). Benzene in blood and phenolin urine in monitoring benzene exposure in industry. Am. J. Ind. Med. 2, 119-123.

Carson, B.L., Ellis, H.V., and McCann, J.L. (1986). Toxicology and BiologicalMonitoring of Metals in Humans. Lewis Publishers, Chelsea, Michigan.

Cralley, L.J. (Ed.) (1985). Patty's Industrial Hygiene and Toxicology. Theory andRationale of Industrial Hygiene Practice-Biological Responses, Vol. 3B, secondedition. John Wiley and Sons, New York. 753 pages.

Droz, P.O. and Guillemin, M.P. (1983). Human styrene exposure. V. Developmentof a model for biological monitoring. Int. Arch. Occup. Environ. Health 53,19-36.

Engstrom, K. (1984). Styrene. In Aitio, A., Riihimaki, V., and Vainio, H. (Eds.)Biological Monitoring and Surveillance of Workers Exposed to Chemicals.Hemisphere, Washington, D.C.

Engstrom, K., Husman, K., and Rantanen, J. (1976). Measurement of toluene andxylene metabolites by gas chromatography. Int. Arch. Occup. Environ. Health36, 153-160.

Engstrom, K., Husman, K., PfaffIi, P., and Riihimaki, V. (1978). Evaluation ofoccupational exposure to xylene by blood, exhaled air and urine analysis. Scand.1. Work Environ. Health4, 114-121.

Fishbeck, W.T., Langner, R.R., and Kociba, R.K. (1975). Elevated urinary phenol

330 METHODS FOR ASSESSING_EXPOSURE OF HUMAN AND NON-HUMAN BIOTA

levels not related to benzene exposure. Am. Ind. Hyg. Assoc. J. 36, 820--824.Gad.El Karim, M.M., Sadagopa Ramanujam, V.M., and Legator, M.S. (1985).

trans, - trans-Muconic acid, an open-chain urinary metabolite of benzene in mice.Quantification by high-pressure liquid chromatography. Xenobiotica 15, 211-220.

Greenberg, R.A., Haley, N.J., Etzel, R.A., and Loda, F.A. (1984). Measuring theexposure of infants to tobacco smoke. New Engl. J. Med. 310, 1075-1078.

Guillemin, P.M. and Bauer, D. (1979). Human exposure to styrene. Part 3:Elimination kinetics of urinary mandelic and phenylglyoxylic acids after singleexperimental exposure. Int. Arch. Occup. Environ. Health 44, 249-263.

Haley, N.J., Axelrad, eM., and Tilton, K.A. (1983). Validation of self-reportedsmoking behaviour: Biochemical analyses of cotinine and thiocyanate. Am. J.Public Health 73, 1204-1207.

Hill, P.and Marquardt, H. (1980). Plasma and urine changes after smoking differentbrands of cigarettes. Clin. Pharmacol. Ther. 27, 652-658.

Isaac, P.F. and Rand, M:J. (1972). Cigarette smoking and plasma levels of nicotine.Nature 236, 308-310.

Jarvis, M., Tunstall-Pedoe, H., Feyerband, e, Vesey, e, and Salloojee, Y. (1984).Biochemical markers of smoke absorption and self reported exposure to passivesmoking. J. Epidemiol. Commun. Health 38, 335-339.

Jensen, A.A. (1983). Chemical contaminants in human milk. Residue Reviews 89,1-128.

Langone, J.J., Gjika, H.B., and Van Vunakis, H. (1973). Nicotine and itsmetabolites: Radioimmunoassays for nicotine and cotinine. Biochemistry 12,5025-5030.

Lauwerys, R. (1979). Human biological monitoring of industrial chemicals. 1.Benzene. Commission of the European Communities, Luxembourg.

Linch, L. (1974). f3iological Monitoring for Industrial Chemical Exposure Control.Chemical Rubber Company, Cleveland, Ohio.

Mikulski, 0.1., Viglusz, R., Bublewska, A., and Uselis, J. (1972). Investigation ofexposure of ships' painters to organic solvents. Br. J. Ind. Med. 29, 450--453.

Ogata, M., Tomokuni, K., and Takatsuka, Y. (1970). Urinary excretion of hippuricacid and m- or p-methylhippuric acid in the urine of persons exposed to vapoursof toluene and m- or p-xylene as a test of exposure. Br. J. Ind. Med. 27, 43--50.

Rosenberg, J., Benowitz, N.L., Jacob, P., and Wilson, K.M. (1980). Dispositionkinetics and effects of intravenous nicotine. Clin. Pharmacol. Ther. 28, 517-522.

Sedivec, V. and Flek, J. (1976). Exposure test for xylenes. Int. Arch. Occup.Environ. Health 37, 219-232.

Senczuk, W. and Orlowski, J. (1978). Absorption of m-xylene vapours through therespiratory tract and excretion of m-methylhippuric acid in urine. Br. 1. Ind.Med. 35, 50--55.

Sheldon, L., Umana, M., Bursey, W., Gutknecht, W., Handy, R., Hyldburg, P.,Michael, L., Moseley, A., Raymer, J., Smith, D., Sparacino, e, and Warner,M. (1986). Biological monitoring techniques for human exposure to industrialchemicals. Noyes Publications, Park Ridge, New Jersey.

Slorach, S.A. and Vaz, R. (1983). Assessment of Human Exposure to SelectedOrganochlorine Compounds through Biological Monitoring. Swedish NationalFood Administration, Uppsala, Sweden.

van Haaften, S.B. and Sie, S.T. (1965). The measurement of phenol in urine bygas chromatography as a check on benzene exposure. Am. Ind. Hyg. Assoc. J.26, 52-58.